Background

I spent fifteen years keeping systems up: servers, directories, storage, backups, and users to help in three languages. That school does not teach theory: it shows how IT stays up day after day, why it goes down, and what every hour of downtime costs.

Data became the core of it: designing and operating pipelines, warehouses and integrations on Google Cloud and Azure, with Airflow, dbt, BigQuery, PostgreSQL and Terraform. From research data to financial feeds, the same standard applies: documented, tested processing that runs overnight without waking anyone.

To that I add the implementation of AI agents, in production: data extraction from document feeds with versioned prompts, code-review and testing agents, assistants that answer with their sources. Always within the same frame: limited tools, an action log, human validation, and sovereign infrastructure, from self-hosted open models to frontier models served from the Zurich cloud regions.

The through-line does not change: reliable systems, documented, explained in plain language. That is what I bring to SMEs and organisations in French-speaking Switzerland: the rigour of large platforms, at the scale of your organisation, and IT that stays in your hands.